Os codificadores preditivos de voz por analise-mediante-síntese vem sendo amplamente aplicados em telefonia móvel celular e em telecomunicações sigilosas. A predição linear do sinal de voz e as técnicas de análise-mediante-síntese são apresentadas de forma a relacionar algumas características perceptivas da audição humana as técnicas e parâmetros usados no processamento de sinais. Esta classe de codificadores e descrita no contexto do codificador preditivo excitado por códigos. Estruturas especiais do codificador tais como livros de códigos adaptativos, esparsos e definidos por base vetorial são abordadas bem como melhoramentos de processamento tais quais as buscas com ortogonalidade. Propõe-se um novo codificador, o codificador preditivo linear com excitação decomposta em vetores singulares, que complementa uma representação recentemente anunciada da excitação da voz com buscas em livros de códigos adaptativos. Os resultados de um estudo de codificadores principais desta classe são apresentados. A analise comparativa baseia-se em medidas objetivas temporais e espectrais. Um estudo suplementar de seleção espectral das características da excitação e de quantização do conjunto completo de parâmetros do codificador proposto revelou resultados interessantes sobre a representação espectral adaptativa e sobre a sensibilidade a quantização das características da excitação. / Analysis-by-synthesis linear predictive speech coders are widely applied in mobile and secure telecommunications. Linear prediction of speech signals and analysis-by-synthesis techniques are presented so that some perceptual features of human hearing may be related to signal processing techniques and parameters. The basic operation of this class of coders is described in the framework of the code-excited predictive coder. Special coder structures such as adaptive, sparse and vector-basis codebooks are introduced as well as processing enhancements such as orthogonal searches. A recently introduced representation of voice excitation is complemented by adaptive codebook searches to give rise to the new proposed coder, the singular-vector-decomposed excitation linear predictive coder. The sults of a study of some important coders in this class is present. The coders are compared on the basis of waveform and spectral objective distortion measures. A further study of spectral selection of excitation features, and quantization of the whole set of parameters is performed on the proposed coder. Some interesting results are described concerning the adaptive spectral representation and the sensitivity to quantization of the excitation features.

In this thesis, the red hot topic anomaly detection is studied, which is a subtopic in machine learning. The company, Procera Networks, supports several broadband companies with IT-solutions and would like to detected errors in these systems automatically. This thesis investigates and devises methods and algorithms for detecting interesting events in diagnostics data. Events of interest include: short-term deviations (a deviating point), long-term deviations (a distinct trend) and other unexpected deviations. Three models are analyzed, namely Linear Predictive Coding, Sparse Linear Prediction and Wavelet Transformation. The final outcome is determined by the gap to certain thresholds. These thresholds are customized to fit the model as well as possible. / I den här rapporten kommer det glödheta området anomalidetektering studeras, vilket tillhör ämnet Machine Learning. Företaget där arbetet utfördes på heter Procera Networks och jobbar med IT-lösningar inom bredband till andra företag. Procera önskar att kunna upptäcka fel hos kunderna i dessa system automatiskt. I det här projektet kommer olika metoder för att hitta intressanta företeelser i datatraffiken att genomföras och forskas kring. De mest intressanta företeelserna är framfärallt snabba avvikelser (avvikande punkt) och färändringar äver tid (trender) men också andra oväntade mänster. Tre modeller har analyserats, nämligen Linear Predictive Coding, Sparse Linear Prediction och Wavelet Transform. Det slutgiltiga resultatet från modellerna är grundat på en speciell träskel som är skapad fär att ge ett så bra resultat som mäjligt till den undersäkta modellen..

<p>A common problem of hearing aids is whistling caused by feedback from the loudspeaker back to the microphone. A method of reducing the negative effects, caused by the feedback, is called feedback cancellation. A variant of feedback cancellation uses a probe signal, which is applied to the speaker of the hearing aid and is used to continuously estimate the feedback. Oticon A/S has suggested a master's thesis with the purpose of designing and evaluating an algorithm generating a probe signal for feedback cancellation systems. The challenge was to find an inaudible probe signal with as much energy as possible. </p><p>Two approaches have been investigated for generating a probe signal. In the first approach the psychoacoustic principle of masking was used to estimate how much noise that could be added to a signal without being heard. Psychoacoustic models, including masking, are used in MPEG (Moving Pictures Expert Group) audio coding and one of these models has been examined in the thesis. In the second approach a standard LPC (Linear Prediction Coding) algorithm was used. In both the MPEG and the LPC approach, warped signal processing has been utilized improving the methods. </p><p>A listening test was performed, evaluating the methods generating the probe signal. The purpose of the test was to determine whether the noise, generated using the MPEG and LPC approach, was inaudible. A hearing aid system with feedback cancellation, using the probe signal, was also simulated. The listening test showed that the noise (probe signal) had to be lowered, much more than expected, to be inaudible. As a consequence, shown in the simulations, the feedback cancellation system, using the probe signal, had trouble identifying the feedback of the hearing aid.</p>

The expected equity premium is an important parameter in many financial models, especially within portfolio optimization. A good forecast of the future equity premium is therefore of great interest. In this thesis we seek to forecast the equity premium, use it in portfolio optimization and then give evidence on how sensitive the results are to estimation errors and how the impact of these can be minimized. Linear prediction models are commonly used by practitioners to forecast the expected equity premium, this with mixed results. To only choose the model that performs the best in-sample for forecasting, does not take model uncertainty into account. Our approach is to still use linear prediction models, but also taking model uncertainty into consideration by applying Bayesian model averaging. The predictions are used in the optimization of a portfolio with risky assets to investigate how sensitive portfolio optimization is to estimation errors in the mean vector and covariance matrix. This is performed by using a Monte Carlo based heuristic called portfolio resampling. The results show that the predictive ability of linear models is not substantially improved by taking model uncertainty into consideration. This could mean that the main problem with linear models is not model uncertainty, but rather too low predictive ability. However, we find that our approach gives better forecasts than just using the historical average as an estimate. Furthermore, we find some predictive ability in the the GDP, the short term spread and the volatility for the five years to come. Portfolio resampling proves to be useful when the input parameters in a portfolio optimization problem is suffering from vast uncertainty.

A common problem of hearing aids is whistling caused by feedback from the loudspeaker back to the microphone. A method of reducing the negative effects, caused by the feedback, is called feedback cancellation. A variant of feedback cancellation uses a probe signal, which is applied to the speaker of the hearing aid and is used to continuously estimate the feedback. Oticon A/S has suggested a master's thesis with the purpose of designing and evaluating an algorithm generating a probe signal for feedback cancellation systems. The challenge was to find an inaudible probe signal with as much energy as possible. Two approaches have been investigated for generating a probe signal. In the first approach the psychoacoustic principle of masking was used to estimate how much noise that could be added to a signal without being heard. Psychoacoustic models, including masking, are used in MPEG (Moving Pictures Expert Group) audio coding and one of these models has been examined in the thesis. In the second approach a standard LPC (Linear Prediction Coding) algorithm was used. In both the MPEG and the LPC approach, warped signal processing has been utilized improving the methods. A listening test was performed, evaluating the methods generating the probe signal. The purpose of the test was to determine whether the noise, generated using the MPEG and LPC approach, was inaudible. A hearing aid system with feedback cancellation, using the probe signal, was also simulated. The listening test showed that the noise (probe signal) had to be lowered, much more than expected, to be inaudible. As a consequence, shown in the simulations, the feedback cancellation system, using the probe signal, had trouble identifying the feedback of the hearing aid.

<p>The expected equity premium is an important parameter in many financial models, especially within portfolio optimization. A good forecast of the future equity premium is therefore of great interest. In this thesis we seek to forecast the equity premium, use it in portfolio optimization and then give evidence on how sensitive the results are to estimation errors and how the impact of these can be minimized.</p><p>Linear prediction models are commonly used by practitioners to forecast the expected equity premium, this with mixed results. To only choose the model that performs the best in-sample for forecasting, does not take model uncertainty into account. Our approach is to still use linear prediction models, but also taking model uncertainty into consideration by applying Bayesian model averaging. The predictions are used in the optimization of a portfolio with risky assets to investigate how sensitive portfolio optimization is to estimation errors in the mean vector and covariance matrix. This is performed by using a Monte Carlo based heuristic called portfolio resampling.</p><p>The results show that the predictive ability of linear models is not substantially improved by taking model uncertainty into consideration. This could mean that the main problem with linear models is not model uncertainty, but rather too low predictive ability. However, we find that our approach gives better forecasts than just using the historical average as an estimate. Furthermore, we find some predictive ability in the the GDP, the short term spread and the volatility for the five years to come. Portfolio resampling proves to be useful when the input parameters in a portfolio optimization problem is suffering from vast uncertainty. </p>

Accurate methods for glottal feature extraction include the use of high-speed video imaging (HSVI). There have been previous attempts to extract these features with the acoustic recording. However, none of these methods compare their results with an objective method, such as HSVI. This thesis tests these acoustic methods against a large diverse population of 46 subjects. Two previously studied acoustic methods, as well as one introduced in this thesis, were compared against two video methods, area and displacement for open quotient (OQ) estimation. The area comparison proved to be somewhat ambiguous and challenging due to thresholding effects. The displacement comparison, which is based on glottal edge tracking, proved to be a more robust comparison method than the area. The first acoustic methods OQ estimate had a relatively small average error of 8.90% and the second method had a relatively large average error of -59.05% compared to the displacement OQ. The newly proposed method had a relatively small error of -13.75% when compared to the displacements OQ. There was some success even though there was relatively high error with the acoustic methods, however, they may be utilized to augment the features collected by HSVI for a more accurate glottal feature estimation.

Digital audio is increasingly becoming more and more a part of our daily lives. Unfortunately, the excessive bitrate associated with the raw digital signal makes it an extremely expensive representation. Applications such as digital audio broadcasting, high definition television, and internet audio, require high quality audio at low bitrates. The field of audio coding addresses this important issue of reducing the bitrate of digital audio, while maintaining a high perceptual quality. Developing an efficient audio coder requires a detailed analysis of the audio signals themselves. It is important to find a representation that can concisely model any general audio signal. In this thesis, we propose two new high quality audio coders based on two different audio representations - the sinusoidal-wavelet representation, and the warped linear predictive coding (WLPC)-wavelet representation. In addition to high quality coding, it is also important for audio coders to be flexible in their application. With the increasing popularity of internet audio, it is advantageous for audio coders to address issues related to real-time audio delivery. The issue of bitstream scalability has been targeted in this thesis, and therefore, a third audio coder capable of bitstream scalability is also proposed. The performance of each of the proposed coders was evaluated by comparisons with the MPEG layer III coder. The first coder proposed is based on a hybrid sinusoidal-wavelet representation. This assumes that each frame of audio can be modelled as a sum of sinusoids plus a noisy residual. The discrete wavelet transform (DWT) is used to decompose the residual into subbands that approximate the critical bands of human hearing. A perceptually derived bit allocation algorithm is then used to minimise the audible distortions introduced from quantising the DWT coefficients. Listening tests showed that the coder delivers near-transparent quality for a range of critical audio signals at G4 kbps. It also outperforms the MPEG layer III coder operating at this same bitrate. This coder, however, is only useful for high quality coding, and is difficult to scale to operate at lower rates. The second coder proposed is based on a hybrid WLPC-wavelet representation. In this approach, the spectrum of the audio signal is estimated by an all pole filter using warped linear prediction (WLP). WLP operates on a warped frequency domain, where the resolution can be adjusted to approximate that of the human auditory system. This makes the inherent noise shaping of the synthesis filter even more suited to audio coding. The excitation to this filter is transformed using the DWT and perceptually encoded. Listening tests showed that near-transparent coding is achieved at G4 kbps. The coder was also found to be slightly superior to the MPEG layer III coder operating at this same bitrate. The third proposed coder is similar to the previous WLPC-wavelet coder, but modified to achieve bitstream scalability. A noise model for high frequency components is included to keep the overall bitrate low, and a two stage quantisation scheme for the DWT coefficients is implemented. The first stage uses fixed rate scalar and vector quantisation to provide a coarse approximation of the coefficients. This allows for low bitrate, low quality versions of the input signal to be embedded in the overall bitstream. The second stage of quantisation adds detail to the coefficients, and hence, enhances the quality of the output signal. Listening tests showed that signal quality gracefully improves as the bitrate increases from 16 kbps to SO kbps. This coder has a performance that is comparable to the MPEG layer III coder operating at a similar (but fixed) bitrate.

A major objective in speech coding is to represent speech with as few bits as possible. Usual transmission parameters include auto regressive parameters, pitch parameters, excitation signals and excitation gains. The pitch predictor makes these coders sensitive to channel errors. Aiming for robustness to channel errors, we do not use pitch prediction and compensate for its lack with a better representation of the excitation signal. We propose a new speech coding approach, Vector Sum Excited Cascaded Linear Prediction (VSECLP), based on code excited linear prediction. We implement forward linear prediction using five cascaded second order sections - parameterized in terms of line spectral frequency - in place of the conventional tenth order filter. The line spectral frequency parameters estimated by the Direct Line Spectral Frequency (DLSF) adaptation algorithm are closer to the true values than those estimated by the Cascaded Recursive Least Squares - Subsection algorithm. A simplified version of DLSF is proposed to further reduce computational complexity. Split vector quantization is used to quantize the line spectral frequency parameters and vector sum codebooks to quantize the excitation signals. The effect on reconstructed speech quality and transmission rate, of an increased number of bits and differently split combinations, is analyzed by testing VSECLP on the TIMIT database. The quantization of the excitation vectors using the discrete cosine transform resulted in segmental signal to noise ratio of 4 dB at 20.95 kbps, whereas the same quality was obtained at 9.6 kbps using vector sum codebooks. / Master of Science

Este trabalho apresenta a pesquisa sobre a extração de parâmetros característicos de sinais acústicos para fins de detecção automática de vazamento de água em tubulações enterradas. Os sinais acústicos foram adquiridos com o auxílio de um geofone eletrônico e também catalogados por técnicos especialistas em detecção acústica. De todos os sinais foram extraídos os modelos de predição linear perceptual de várias ordens, determinando-se como melhor a ordem 2. A partir de um conjunto de modelos de referência de sinais de vazamento, a distância média de Itakura dos outros modelos em relação a estas referências foram calculadas. Em conjunto com estas distâncias, quatro características espectrais são também extraídas do sinal a fim de compor o vetor de parâmetros característicos do sinal. Parte destes vetores de parâmetros característicos são utilizados para treinar o classificador de máquina de vetores de suporte. O restante dos dados são, então, submetidos a este classificador que obteve a taxa de acerto de classificação em torno de 93%. Experimentos anteriores, utilizando modelos de predição linear, de ordem 10, obtiveram uma taxa de acerto em torno de 82%. Isso demonstra que estes novos parâmetros característicos propostos alcançam os objetivos deste trabalho, que são algoritmos com melhor taxa de acerto na detecção de vazamentos. / This work presents a research about feature extraction of acoustic signals for detection of water leak in buried pipes. Acoustic signals were acquired by means of an electronic geophone and also labeled by technicians specialized in acoustic water leak detection. For every signals, its linear predictive model was estimated for a range of prediction orders, concluding for the best order 2. Out of this group of models, some leaky ones are used as reference for calculating the Itakura mean distance with respect to the other models. Completing this measure, four spectral features are extracted to compose the signal feature vector. Some of these vectors were used to train a support vector machine to be used as a classifier. The remaining ones were used to evaluate the classification. The resulting accuracy rate achieved is around 93%. Earlier experiments, which use linear prediction of order 10 had an accuracy rate around 82%. This shows that this novel proposal of feature vector achieves the main goal of this research, which is the increase in the leak detection accuracy rate.